Fix broken main branch by providing correct getBitSet implementation

[BewareMyPower]

Motivation

Currently the main branch is broken by the concurrent merge of #153 and #151. It's because when a batched message id is constructed from deserialization, there is no getBitSet implementation of the internal acker.

Modifications

Add a bool parameter to MessageIdImpl::getBitSet to indicate whether the message ID is batched. The logic is similar with

https://github.com/apache/pulsar/blob/299bd70fdfa023768e94a8ee4347d39337b6cbd4/pulsar-client/src/main/java/org/apache/pulsar/client/impl/PersistentAcknowledgmentsGroupingTracker.java#L325-L327

and

https://github.com/apache/pulsar/blob/299bd70fdfa023768e94a8ee4347d39337b6cbd4/pulsar-client/src/main/java/org/apache/pulsar/client/impl/PersistentAcknowledgmentsGroupingTracker.java#L345-L347

Add a testMessageIdFromBuild to test the acknowledgment for a message ID without an acker could succeed for a consumer that enables batch index ACK.

TODO

In future, apache/pulsar#19031 might be migrated into the C++ client to fix the consumer that disables batch index ACK.

Documentation

• doc-required
  (Your PR needs to update docs and you will update later)

• doc-not-needed
  (Please explain why)

• doc
  (Your PR contains doc changes)

• doc-complete
  (Docs have been already added)

[shibd]

I have a question:

It's because when a batched message id is constructed from deserialization, there is no getBitSet implementation of the internal acker.

Is the code related to this passage here?

pulsar-client-cpp/lib/MessageIdBuilder.cc

Lines 45 to 47 in d03ff20

	if (impl_->batchIndex_ >= 0 && impl_->batchSize_ > 0) {
	return MessageId{std::make_shared<BatchedMessageIdImpl>(*impl_, BatchMessageAckerImpl::create(0))};
	} else {

Is setting bitSize equal to 0 for lazy initialization? To save memory?

If bitSite is initialized here, subsequent operations such as ack can reuse the logic here. Wouldn't it be better?

pulsar-client-cpp/lib/BatchedMessageIdImpl.h

Lines 38 to 40 in f860566

	bool ackIndividual(int32_t batchIndex) const { return acker_->ackIndividual(batchIndex); }
	bool ackCumulative(int32_t batchIndex) const { return acker_->ackCumulative(batchIndex); }

[shibd]

The modification of this class to have solved the compilation error. Can we first open a new PR to merge it?

[BewareMyPower]

I will modify this PR after #177 is merged.